The Association Between Cadmium Exposure and Osteoporosis: A Longitudinal Study and Predictive Model in a Chinese Female Population

Objective: The association between cadmium exposure and osteoporosis has been rarely reported in longitudinal studies. In this study, we investigated the association between osteoporosis and cadmium exposure and developed predictive models in women in a longitudinal cohort. Materials and Methods: In total, 488 women living in southeastern China were included at baseline (1998). Cadmium in blood (BCd) and urine (UCd) and also renal dysfunction biomarkers and bone mineral density (BMD) were determined both at baseline and follow-up. A total of 307 subjects were finally included after excluding subjects that did not have exposure or effect biomarkers. Osteoporosis was defined based on T score ≤ −2.5. Multiple linear regression and multivariate logistic analysis were used to show the association between baseline data and follow-up osteoporosis. Based on the identified associated factors, nomograms were developed to graphically calculate the individual risk of osteoporosis. Results: The baseline BMD in subjects with osteoporosis was significantly lower than that in subjects without osteoporosis (0.59 vs. 0.71 g/cm(2), p < 0.05). The prevalence of low bone mass at baseline was higher in subjects with osteoporosis than in those without osteoporosis (23.5 vs. 7.2%, p = 0.001). Logistic regression analysis demonstrated that age [odds ratio (OR) = 1.21, 95% confidence interval (CI): 1.16–1.27], UCd (OR = 1.03, 95% CI: 1.002–1.06) and the presence of low BMD (OR = 3.84, 95% CI: 1.49–9.89) were independent risk factors for osteoporosis. For those subjects with normal baseline BMD, age, UCd, and baseline BMD were also independent risk factors for osteoporosis. The OR value was 1.16 (95% CI: 1.10–1.22) for age, 2.27 (95% CI: 1.03–4.99) for UCd > 10 μg/g creatinine, and 0.39 (95% CI: 0.21–0.72) for BMD(baseline). We developed two nomograms to predict the risk of osteoporosis. The area under the curve was 0.88 (95% CI: 0.84–0.92) for total population and was 0.88 (95% CI: 0.84–0.92) for subjects with normal baseline BMD, respectively. Conclusion: Baseline age, UCd, and BMD were associated with follow-up osteoporosis in women. Nomograms showed good performance in predicting the risk of osteoporosis.